Ground Issues on MK-5=>MK-6e Migration

We are in the middle of a MK-5 to MK6e migration and have encountered an issue of grounds in the new system.

Whenever a core is powered off (with the toggle switch in <PD>), a split occurs on the DC system. With all cores powered up there is no split, with one core powered down there is a split of 10-15V, with all 3 down there is about 70V split.

The initial response from the OEM was that "This is as per the original GE MK-5 design". However, we went to a sister unit that has not been migrated and found that powering off cores in the un-migrated MK-5 panel does not generate splits.

Has anyone ran into this issue before?
Mike, are you seeing the actual split change as noted by using a voltmeter?

Or just seeing the displayed split change on the HMI?
Is this a MKV migration? Replacement of the MKV cores with MK6E controllers, but still using the original TB's?
Or a full rip-replace to full MK6E?

I have seen some changes to the calculated split on a MKVE, or MKV migration when an issue with the power supply or MVRP for an individual core was failing. In this case it indicated a 125VDC ground when the power supply for the <S> core was failing. There are jumpers on each MVRP that need to be in place, and the Berg jumper at the PD core is supposed to be removed during the upgrade, if this is a MKVE or migration.

Hope this helps.